Lidar plays an essential role in monitoring the vertical variation of atmospheric aerosols. However, due to the limited information that lidar measurements provide, ill-posedness still remains a big challenge in quantitative lidar remote sensing. In this study, we describe the Basic algOrithm for REtrieval of Aerosol with Lidar (BOREAL), which is based on maximum likelihood estimation (MLE), and retrieve aerosol microphysical properties from extinction and backscattering measurements of multi-wavelength Mie–Raman lidar systems. The algorithm utilizes different types of a priori constraints to better constrain the solution space and suppress the influence of the ill-posedness. Sensitivity test demonstrates that BOREAL could retrieve particle volume size distribution (VSD), total volume concentration (V_t), effective radius (R_eff), and complex refractive index (CRI = n − ik) of simulated aerosol models with satisfying accuracy. The application of the algorithm to real aerosol events measured by LIlle Lidar AtmosphereS (LILAS) shows it is able to realize fast and reliable retrievals of different aerosol scenarios (dust, aged-transported smoke, and urban aerosols) with almost uniform and simple pre-settings. Furthermore, the algorithmic principle allows BOREAL to incorporate measurements with different and non-linearly related errors to the retrieved parameters, which makes it a flexible and generalized algorithm for lidar retrieval.

Mineral dust is a key atmospheric aerosol agent that impacts the radiation budget and plays a significant role in cloud formation. However, studies on retrieving height-resolved microphysical properties of dust aerosols, which are crucial for understanding dust evolution, transport processes, and radiative effects, from lidar measurements are still insufficient. Here, we retrieve dust aerosol microphysical properties, including the volume size distribution (VSD), total volume concentration (V_t), effective radius (r_eff), complex refractive index (CRI), and single-scattering albedo (SSA), from spectral extinction (α), backscattering (β), and depolarization (δ) lidar measurements. We evaluate the performance of three particle scattering models, namely the spherical, spheroidal, and irregular–hexahedral (IH) models, in terms of mimicking dust optical properties and deriving retrieval results when different measurement combinations are inverted. Both simulations and inversions of real lidar measurements confirm the superiority of the IH model and the significance of spectral depolarization measurements to improve the retrieval accuracy. An increase in discrepancy in depolarization ratio produced by the IH and spheroid models is observed for r_eff>0.5 µm, resulting in larger retrieval difference between the two non-spherical models after the inclusion of 3δ. Comparisons of the real case retrievals with Aerosol Robotic Network (AERONET) retrievals and previous in situ results indicate relatively smaller r_eff and larger SSA derived from the lidar retrievals. A discussion of the possible reasons is presented.

The accuracy of height‐resolved dust microphysical retrieval from LiDARs has been greatly improved by the recently developed BOREAL (Basic algOrithm for REtrieval of Aerosol with LiDAR) algorithm which describes non‐spherical dust particles with the Irregular‐Hexahedral (IH) model and inverts 3 β (backscattering at 355, 532, and 1,064 nm) + 2 α (extinction at 355 and 532 nm) + 3 δ (depolarization ratio at 355, 532, and 1,064 nm) measurements. However, potential retrieval bias in BOREAL was observed as the dust coarse‐mode radius ( R _c ) increased, probably due to the reduction in measurement sensitivity. In this study, a parameterized‐constrained algorithm (BOREAL‐PC) is proposed and developed to improve the quality of dust microphysical retrieval. It is based on optimal estimation and makes use of historical in situ and laboratory measurements to provide appropriate a priori constraints. Simulations based on synthetic optical data show apparent improvement in retrieval accuracy and validate the error model for BOREAL‐PC's retrievals. However, the retrieval quality declines for particles either too small ( R _c ∼ 1 μm) or too large ( R _c ∼ 6 μm) with respect to the a priori constraint due to the rise of smoothing error. Real case studies show that compared to BOREAL, BOREAL‐PC retrieves dust microphysical properties which are more consistent with the AERONET retrievals. Furthermore, the potential of BOREAL‐PC for dust‐cloud interaction study is demonstrated by comparing the CCN‐ (cloud condensation nuclei) and INP‐ (Ice Nucleating Particle) relevant aerosol parameters retrieved by BOREAL‐PC with the corresponding POLIPHON retrievals and aircraft in situ measurements in a case study.

Mineral dust is a key atmospheric aerosol agent that impacts the radiation budget and plays a significant role in cloud formation. However, studies on retrieving height-resolved microphysical properties of dust aerosols, which are crucial for understanding dust evolution and transport processes, from lidar measurements are still insufficient. Here, we retrieve dust aerosol microphysical properties, including the volume size distribution, volume concentration, effective radius (r_eff), refractive index and single-scattering albedo, from spectral extinction, backscattering and depolarization lidar measurements. We evaluate the performance of three particle scattering models – Sphere, Spheroid and Irregular–Hexahedral (IH) models in terms of mimicking dust optical properties and deriving retrieval results. We also explore the influence of inverting different measurement sets, namely the conventional 3β (backscattering coefficients at 355, 532 and 1064 nm) + 2α (extinction coefficients at 355 and 532 nm) and the expanded 3β + 2α + 3δ (depolarization ratio at 355, 532 and 1064 nm) measurements, on the retrieval. Both simulations and inversions of real lidar measurements show that it is necessary to use non-spherical models and incorporate 3δ measurements to improve the retrieval accuracy. An increase of discrepancy in depolarization ratio produced by the IH and Spheroid models is observed for r_eff > 0.5 μm, resulting in larger retrieval difference between the two non-spherical models after the inclusion of 3δ. The study demonstrates the prospect of retrieving height-resolved dust microphysical properties from lidar measurements, as well as potential limitations of the prevailing scattering models in simulating particle backscattering properties.

Mineral dust is a key atmospheric aerosol agent that impacts the radiation budget and plays a significant role in cloud formation. However, studies on retrieving height-resolved microphysical properties of dust aerosols, which are crucial for understanding dust evolution and transport processes, from lidar measurements are still insufficient. Here, we retrieve dust aerosol microphysical properties, including the volume size distribution, volume concentration, effective radius (r_eff), refractive index and single-scattering albedo, from spectral extinction, backscattering and depolarization lidar measurements. We evaluate the performance of three particle scattering models – Sphere, Spheroid and Irregular–Hexahedral (IH) models in terms of mimicking dust optical properties and deriving retrieval results. We also explore the influence of inverting different measurement sets, namely the conventional 3β (backscattering coefficients at 355, 532 and 1064 nm) + 2α (extinction coefficients at 355 and 532 nm) and the expanded 3β + 2α + 3δ (depolarization ratio at 355, 532 and 1064 nm) measurements, on the retrieval. Both simulations and inversions of real lidar measurements show that it is necessary to use non-spherical models and incorporate 3δ measurements to improve the retrieval accuracy. An increase of discrepancy in depolarization ratio produced by the IH and Spheroid models is observed for r_eff > 0.5 μm, resulting in larger retrieval difference between the two non-spherical models after the inclusion of 3δ. The study demonstrates the prospect of retrieving height-resolved dust microphysical properties from lidar measurements, as well as potential limitations of the prevailing scattering models in simulating particle backscattering properties.

Mineral dust is a key atmospheric aerosol agent that impacts the radiation budget and plays a significant role in cloud formation. However, studies on retrieving height-resolved microphysical properties of dust aerosols, which are crucial for understanding dust evolution and transport processes, from lidar measurements are still insufficient. Here, we retrieve dust aerosol microphysical properties, including the volume size distribution, volume concentration, effective radius (r_eff), refractive index and single-scattering albedo, from spectral extinction, backscattering and depolarization lidar measurements. We evaluate the performance of three particle scattering models – Sphere, Spheroid and Irregular–Hexahedral (IH) models in terms of mimicking dust optical properties and deriving retrieval results. We also explore the influence of inverting different measurement sets, namely the conventional 3β (backscattering coefficients at 355, 532 and 1064 nm) + 2α (extinction coefficients at 355 and 532 nm) and the expanded 3β + 2α + 3δ (depolarization ratio at 355, 532 and 1064 nm) measurements, on the retrieval. Both simulations and inversions of real lidar measurements show that it is necessary to use non-spherical models and incorporate 3δ measurements to improve the retrieval accuracy. An increase of discrepancy in depolarization ratio produced by the IH and Spheroid models is observed for r_eff > 0.5 μm, resulting in larger retrieval difference between the two non-spherical models after the inclusion of 3δ. The study demonstrates the prospect of retrieving height-resolved dust microphysical properties from lidar measurements, as well as potential limitations of the prevailing scattering models in simulating particle backscattering properties.